HDL, both in humans and animal models, has been shown to be atheroprotective. Human HDL is heterogeneous and consists of two major subclasses, HDL2 and HDL3. Epidemiological studies suggest that HDL2 is more atheroprotective than HDL3. Mice and pigs have a monophasic HDL profile. In mice the HDL is close in size and density to human HDL2, and in pigs it is similar to HDL3. The primary goals of this proposal are to generate a mouse model in which HDL2 or HDL3 are the predominant HDL subclass and to test the atheroprotective effects of HDL2 and HDL3 in a well characterized murine atherosclerotic model. In the first specific aim we will use site-directed mutagenesis to make human apoA-I mutants or human/mouse and human/pig chimeric apoA-I and determine if they demonstrate a preferential association in vitro with mature human HDL2 or HDL3, respectively. Our goal is to generate a protein with minimum sequence differences from human apoA-I. Based on our preliminary data we will initially focus on the interhelical turn between helices 7/8. This turn region will be substituted with human interhelical turns containing proline residues or with interhelical turns from mouse or pig. Our second aim will be to demonstrate that selected apoAI mutants characterized in specific aim 1 can generate HDL2 and HDL3 in vivo in apoA-I deficient mice by adenoviral mediated gene transfer. The, third specific aim will test the efficacy of the engineered HDL2 and HDL3 in protecting against the development of atherosclerosis in human apoB transgenic mice expressing, as knockin genes, the apoA-I proteins that best form HDL2 and HIDL3. The final specific aim will examine how the mutants or the engineered HDLs generated in vivo interact with enzymes and receptors that are involved in HDL remodeling and cholesterol efflux.